The heat detection, done without any technologies could be challenging because of small number of workers and high number of animals in dairy farms. The technology that is most commonly used for heat detection in dairy farms is step counters of pedometers, assigned individually to every animal.

The pedometers are showing not only increased activity that usually means that cow is in heat, but also decreased activity that usually leads to conclusions that animal has leg or claw problems or health disorders. Use of step counters positively impacts the length of calving interval and number of calves obtained from one cow, not only per year, but also per whole cows life.

The introduction of pedometers in dairy holdings usually leads to decreased number of missed heats, properly assigned service time and early detection of lameness cases. In different local studies by taking local farm animals, it was determined that use of pedometers also showed positive impact on dairy cow longevity.

In the may Latvian dairy farms step counters are introduced as addition to automatic milking systems and herd management systems to collect data about animal activity and location in cowshed.

In farm, where step counters were introduced in year 2010 the number of missed heats lowered by 20% and number of services per conception were reduced by 0.4 times. Average calving interval was shortened by 64 days. The number of early detections of lameness cases increased by 35%.

In the fast pace of life and events, it is important to find out all information about the animal and its production as soon as possible. Many of available dairy sensors give farmer information about the condition of dairy animal, but there still are information that could be crucial in determination of the health, wellbeing and feeding condition of dairy animals. The small milk laboratories in automatic and parlor milking systems are used to obtain data about milk composition and quality (milk fat and protein content, lactose and milk urea content as well as somatic cell count in milk).

From obtained data it is possible to determine cows metabolic state (milk fat and protein ratio), feed efficiency (milk urea content, milk fat content), udder health (somatic cell count) etc. It also allows to separate the milk with poor quality from sellable milk (the electroconductivity sensors in milking systems).

By introducing the small milk sample laboratories in the milking system, it became possible to faster evaluate and improve feed ration, treat udder inflammation, monitor cows productivity etc.

By introducing small milk sample laboratories in farm, the most significant impact was on the metabolic disorder count (decreased by 25%) and udder inflammation monitoring. In average one case of mastitis were spotted 1.8 days faster than appeared first clinical signs.

In modern day agriculture one of most excluded branches is organic dairy farming. In organic dairy farming the largest emphasis is on cow welfare. The common assumption is that in loose housing system with all year access to grazing and walking area is not suitable for the best option to ensure the closest to optimal conditions for cow welfare. Nowadays there are organic farms that brakes that assumption and leads by example that it is possible to grant optimal living conditions for cows, kept and fed according to organic farming rules.

The farm in focus was established in year 2004 with only 12 dairy cows and in year 2018 it has increased to 89 animals with average age 3.4 lactations (in year 2017 in herd was introduced 23 pregnant heifers). In year 2014, with the support of European Union, was built modern cowshed, according to organic farming rules, and in timespan from 2015 to 2018 in farm were introduced different modern technologies. At the end of year 2018 in farm were introduced herd management systems, neck collars with location tags, automatic milking system, semi-automatic climate control system, feed pusher robot and automatic fodder feeding stations.

Throughout the years in farm is observable tendency for milk productivity increase (in year 2018 it was recorded on 8560 kg per one cow) and steady somatic cell count in milk (around 220 – 250 thousand in 1 mL-1). With the introduction of feed pusher, cows tended to be calmer and with that improved feed intake and efficiency.

The heat monitoring and service efficiency rapidly improved after the introduction of herd management system and in year 2018 the average calving interval was 387 days and average number of services per one calving was 1.4.

In modern dairy farming use of herd management systems is almost mandatory, because the amount of collected in-farm data nowadays goes beyond human capacity. Modern herd management systems collect data form different in-farm sensors and assists to farmer with the well thought out reports that helps to manage dairy herd.

The data management system serves as on-farm advisor for farmer and farms workers that suggests about the manipulations that needs to be done (service, removing from group, moving to nursery etc.) based on the all the data, collected from farm animals.

By inputting the information about important events of cows life (service, calving, etc.) it is possible to obtain a proper herd documentation that is suitable for quick redacting and covers all cows life events in quick and easy accessible way.

In local dairy farms that has more than 100 dairy cows, herd management system is not rare phenomenon. In different farms are used different management systems depending on the manufacturer of other in-farm sensors, used in holding.

By introducing herd management systems in farm, in addition to easier manageable herd documentation, there are observable significant decrease of length of calving interval and days open, due to better heat monitoring, as well as decreased somatic cell count in milk and number of health disorders in farm, due to better monitoring of monitorable traits (somatic cell count, number of steps per day, body temperature, etc.)

The effectiveness of ration quality, the state of cows metabolic and overall health shows significant impact on their body condition. In the large groups of animals, it is almost impossible to notice the changes of the body condition (BC) of one cow, but, by introducing modern technologies into cows life, it is possible to have cow BC score measurements several times per day. One of such technologies is cow body condition scoring camera, usually located before milking parlor. This camera is reading data every time when cow enters milking parlor gates (2 – 3 measurements per day). This system allows to follow the BC of each individual animal and spot any deviations from norm. The knowledge about BC score of individual animal and cow group serves as helpful indicator on quality of prepared feed ration. By using the data about cow BC and analyzing the correlation with milk composition data it is possible to evaluate the feeding efficiency as well as the metabolic state of each individual cow and in cow group.

In the farm with 100 dairy cows BC scoring camera were introduced in year 2011. In this time the number of cows with metabolic disorders decreased by 68%, milk productivity increased by 2800 kg (25% increase) and calving interval decreased by 46 days. In the farm BC scoring cameras are used for determination of the feed ration effectiveness and its impact on cows. The feed ration is adapted if there are negative response on the quality or ratio of different feedstuff.

The automatic milking systems or milking robots are one of most introduced technologies in Baltic state dairy farms in last 5 years. They are usually introduced because of lack of capable workforce and because of their positive impact on dairy cow welfare (lowered stress factor impact). The automatic milking system ensures high quality udder disinfection before and after milking as well as standardized milking procedure without any human interaction.

Introduction of milking robots usually related to high levels of stress for animals at the beginning of introduction of the system, but later the stress levels for the cows milked by robot is significantly lower than it is in cow groups in that are milked in milking parlors. The lower stress is usually linked with standardized milking procedure that is repeated in every milking and does not scare animals.

The common practice in dairy farms where are introduced milking robots shows that by doing that the milk productivity increases significantly – after adaptation period the amount of milk obtained in one day increases by 15 – 35%, depending on the in-farm conditions before installation of automatic milking systems.

Introduction of milking robots in dairy farms usually comes with increased quality of obtained milk, because the system fully excludes the possibility for milk to interact with air and the udder disinfection is done more carefully. In farms with milking robots, introduced in last 5 years (from year 2013) the amount of somatic cells in milk had lowered by 50 – 75%.

In modern dairy farming conditions, the need for precise, balanced and continuous feed distribution method becomes crucial. One of methods to ensure the satisfaction of all mentioned needs is automatic feed ration mixers and distributers.

In family owned farm with 160 dairy cows, after long and unsuccessful search for local labor and cost analysis of traditional feed distribution, in year 2010 was introduced automatic feeding system. System in farm is equipped with feed storage stations, ration mixer and distributer, measuring lasers for feed quantity assessments and feed pusher.

The system is programmed to feed each cow group once in four hours that ensures that even in nights on the feeding table will be available fresh feed. In system it is possible to prepare and distribute feed ration individually for each cow group.

One of main advantages of the system in farm was lowered need for manual labor as well as more precise and continuous distribution of feed. The automatic feeding system also shows significant impact on cow welfare, especially with the lowering of stress factors.

With the introduction of automatic feeding station in the farm cow average milk productivity increased by 3750 kg with additional increase of milk fat and protein content. In addition to that in farm decreased the number of cows with different traumas, because of lowered hustling at feeding table.

No less important is the fact that farm owner does not need any additional labor for cow feeding organization. The systems feed storage can store silage for up to 5 days that allows famer to be more flexible with his time management, therefore improving his quality of life.

One of most important factors that affects dairy cow productivity as well as overall body condition is the climate conditions in cowshed, especially the air humidity and temperature levels. In Eastern Europe located farms characterizes with extremely high changes of air temperatures in-between seasons – hot and humid summers (+25 – 300C) and cold and humid winters (-20 - -250C) and in recent cases there are severe changes of environmental temperature during short periods of time. Therefore, in dairy farms, climate control systems are crucial.

In most of modern farms automatic climate control systems are included up until some extent – either it is automatically controlled air ventilators or automatic curtain system or automatic heaters in strategic points of cowshed.

The system reads and analyzes the temperature and average humidity level throughout the cowshed and sends signal to open or close curtains and turn on/off ventilators and heaters. The system equipped in dairy farm (600 dairy cows) also contained automatic daylight saving – it turns lights on or off depending on the outside lighting. This system allows to save additional money for farmer from the lowered electro energy expenses that comes from useless lighting of cowshed.

Nevertheless, some of climate control systems in dairy farms are equipped with water misters next to feeding tables and in waiting area before milking parlor. That allows reduce the negative effects of heat in hot summers.

The automatic climate control in farm showed positive impact on cow welfare and overall wellbeing of cows. With the introduction of cowshed climate control system in farm, the number of cold stress decreased by 75%, but number of heat stress cases decreased by 50%.

In dairy farms with loose housing system with limited human contact it is quite hard to notice every cow in heat and be present at every calving. Therefore, in some dairy farms, more often are used rumen boluses that read data about cow temperature, activity and rumen pH. Data from boluses are collected in computer program that makes analysis and gives best predictions of different cows’ life events.

In family owned farm, in the beginning of year 2018 were introduced 42 rumen boluses. The information about cow daily activity serves as indicator for heat detection, data about body temperature are used to prognose calving and rumen pH results are used to evaluate the quality of feed ration, cows metabolic state and quality of rumination.

The introduction of boluses lowered number of days open (from 148 to 123 days) as well as decreased number of services per conception (from 2.3 to 2.1) mainly due to data, now available for farmer.

The system in farm also gives additional information about cow’s body temperature that can serve as indicator for different inflammatory or metabolic processes in rumen or overall cows body. In past, from farmers perspective, data from boluses showed significant effect on the precision of heat detection and prognosis of the most suitable time for service as well as on the management of cow health condition in farm.

Katri Ling and David Arney

Estonian University of Life Sciences

Alongside recent increases in dairy cow milk yields is a concomitant decrease in the cows’ fertility. The causes of these reductions in fertility may be multiple, but the outcome for the farmer is increased calving intervals and premature culling leading to impaired efficiency and consequent economic losses. Recording milk progesterone can potentially accurately identify the timing of oestrus and pinpoint the optimum time for successful insemination, allowing the farmer to avoid missing heats and reducing the calving interval. A large dairy farm in Estonia, with approximately 400 cows and a high average milk yield of 11,000 kg per year, has been involved in a pilot project with Estonian Livestock Performance Recording Ltd to test progesterone routinely in their cows using the DeLaval Herd Navigator™ system. The farm management are very satisfied with the results from their use of this system, with the calving interval having been reduced from 410 days when the project started to 389 days. In part as a consequence of these findings, four farms have now started to use this system in Estonia.

Annemari Polikarpus and David Arney

As dairy herds get bigger across Europe the ability for farmers to give individual attention to their cows becomes increasingly unfeasible. Identification of health problems may go unnoticed until a late stage of the progress of the disease, when treatment is more costly, less effective and the damage to the animal’s productive potential has already been done. The order in which cows enter the milking parlour can be routinely collected in precision livestock farming. This entrance order is usually quite stable, but if cows are sick or injured they may enter the parlour in a different position to usual in the entry order. Maybe because they are fearful of pain in the parlour or during the milking process, or because of their sickness their position in the social hierarchy is affected. A study in Estonia looked at the effects of impaired on health on the change in the order that cows enter the milking parlour. This was found to be true for cows with mastitis and with metritis. The regular monitoring of milking order, and flagging of changes in this order, could be an effective and low-cost tool in the early identification of the presence of sub-clinical disease in individual cows in large loose-housed dairy farms. While such a system is not currently in place, Delaval and other parlour milking equipment providers do record routinely the ID of cows when they enter the parlour and farmers could use this data source to identify changes in the order and alert them to individual cow problems.

Often technology is proposed as a way to help farmers in order to get more productive. However, often it is not the desire to increase milk yield per cow that motivates farmers to consider a new set of technology such as e.g. a milking robot, yet the need to organize labor more appropriately. When a farmer exceeds a number of livestock (say more than 200) he will be in need for an additional working force. When a farmer is not able to find an adequate candidate with a passion for the field and a long term engagement, this might convince him to consider a robot. Is this, however the right reason? This illustrates how PLF technology also needs to be considered in relation to the larger societal system in which farming takes places: maybe we need to invest in a match-making system between skilled labor and dairy farmers? If a robot makes it possible that the ageing farmer keeps on helping his relatives that have taken his farm, does this automatically mean that this is to be considered as something normal.

What is needed to tackle this problem is a serious research and debate about the relationship between technology and labor efficiency. It needs to be identified how the labor system in a particular regions can be optimized and how the role of technology and sensors can be instrumental to a farming labor experience that is based on the joy of farming life, and not on a coping strategy to keep on maximizing productivity.

Project-action on both European and national levels is much needed.

The use of sensors in dairy farming industry can be seen as a quite recent phenomenon. More and more farmers are now beyond the first generation of activity meters geared at fertility management, moving towards more complex systems that also monitor data relevant for feed strategies and animal welfare. Since a lot of farmers are only just starting to work with these systems, the usability of the data for long term dairy farm decisions is still not at a mature point. In a few years, farmers will however have access to long-term data series, potentially supporting them in more long-term strategic decisions.



Sensor data is used to look at year-by-year results about animal welfare and feed management, but also on time investment on specific hours spent on specific tasks. For instance, what time has been spend on animal welfare management when comparing young stock breeding with the dairy herd. What has been the occurrence of diseases and how much time has been spent on treatment and with which success. Feedback of this kind can be drawn from sensor software and can guide dairy farms in case a strategic decision needs to be taken: for instance, will I keep on doing young stock breeding myself or will I outsource this activity?



Also in relation to feed strategies longer term data series could provide valuable information. What have been crucial changes over the years in terms of feed choices and what has been the impact on both milk yield and profitability. Sensor data can thus provide a kind of accurate feedback on the history of day-to-day decisions adding a new layer to strategic decision making.

The proportion of automatic milking systems (AMS) is continuously increasing throughout Europe. Robotic milking offers great advantages in terms of labour flexibility and data collection, but comes with a steep learning curve. One of the biggest challenges for farmers using AMS, is the high number of false mastitis alerts. So how can dairy farmers distinguish true from false alerts?

Every commercial AMS model uses a different method to generate a mastitis alarm. The most frequently used factors for monitoring mastitis, however, are the electrical conductivity of milk and milk yield measurements. Both can be measured per quarter. Theoretically, visually inspecting every single mastitis alarm in the barn is the best option to detect clinical mastitis – but this would result in an impossible workload.

The first step in controlling mastitis on AMS farms, is checking the computer at least twice a day. The software will automatically generate “attention lists”, highlighting the cows that require the farmer’s attention. Still, not every cow on the attention list has mastitis, and not every cow with mastitis requires an antibiotic treatment. Combining multiple variables (electrical conductivity, milk yield, color alerts, …) with the cow’s known history (days in lactation, milking interval, …) is essential for properly evaluating mastitis alerts. Cows that turn up more than once on the attention list should definitely be kept an eye on.

Blanket dry cow therapy, meaning all cows receive an infusion of antimicrobials in their 4 quarters at dry off, has been standard practice for years. However, with growing concerns for antimicrobial resistance, more farms are shifting towards selective dry cow therapy. This means you are only going to treat the animals that are suspected of having an infection. But how do you select the cows that should get antimicrobials? Use data!

Cows with:

- No clinical mastitis in the previous lactation

- Milk production lower than 15 kg

- Somatic cell count at the 3 latest controls lower than 150.000 (heifers) and 100.000 cell/ml (cow)

… can be selected for dry off without antibiotics. In all other situations, you can take a milk sample for culturing. Good dry off management without antibiotics is possible, but excellent record keeping and hygiene at dry-off is crucial. Possible benefits of drying off without antibiotics include reducing the risk of antibiotic resistance while saving money in the process.

The success of automatic milking systems largely depends on the cow’s own motivation to visit the milking unit several times a day. The number of cows that need to be fetched twice a day should not exceed 5% of the herd. Ideally, the average milking frequency does not drop below 2.4 times per day. But what if it does?

The first thing to do, is to analyse the potential problem. By only looking at herd averages, you run the risk of glossing over potential problems in a specific subset of cows. Low-yielding cows at the end of their lactation do not need to visit the milking robot as often. The high-producing cows in their early lactation, on the other hand, should find their way to the milking unit easily, and reach a milking frequency of 3 times a day – or more.

Ultimately, the cow’s main motive to visit the milking robot is feed. If the number of visits starts to decline, check the concentrate dispenser in the milking robot. Also, palatability is key in “luring” the cows to the milking robot. Even minor changes in the quality or composition of the concentrates or forage at the feed bunk can alter the visiting behaviour of the cows.

But every farm is different, of course. General management factors such as herd size, access to pasture, barn design, type of cow traffic, … all have a huge effect on the overall milking frequency. Sick or lame cows will also be less motivated to go to the milking robot. Make sure your cows are in good health when you notice the visiting frequency starts to decline.

Hanno Jaakson and Priit Karis, Estonian University of Life Sciences

Understanding cows’ body condition score (BCS) is a valuable tool in the early identification of health and feeding problems and also extended periods and extent of negative energy balance (NEB) post partum. This is becoming more of a problem as cow milk yields increase and the ability to match intakes to requirements becomes more difficult to manage. Identifying NEB early in cows at risk is of great benefit to farmers. Manually assessing BCS is time consuming and to do so accurately takes skill and experience, while automatic devices can save time and labour and, if accurate, can give early warning of problems to farmers. A DeLaval automatic BCS device was tested on a 600-cow Holstein-Friesian dairy farm in Estonia. Sixty-six cows were compared using the automatic system and with trained assessors using visual assessment with a classic five-point BCS scale. The two methods gave very close scores, the manual assessment very slightly underscoring compared with the automatic system. This preliminary evaluation shows that the system works well and can be used by farmers to quickly and reliably identify those cows with too low BCS and allow the early management of that cow to prevent the problem from worsening.

What is the real amount of dry-matter-content of the harvested grass, is one of the most asked questions of farmers during harvesting which stays most of the time unanswered. The dry-matter-content is besides protein and feed-value on of the most used characteristics of grass during harvesting in summer and making up the right ration for the herd in the winter. So, even it is important information for a farmer to know, it is still difficult and time-consuming for farmers to get this information by themselves.

There are several sensor-based yield measurements available which can be installed on the silage machines. For example the Pasture Reader. The Pasture Reader measures with ultrasonic technology the height and the density of the grass, this measurement results in the dry-matter-content in kg/ha. The Pasture Reader can be installed on a mower and/or quad. Besides ultrasonic technology also Near-Infra-Red-Spectroscopy (NIRS) is a possibility. NIRS can measure the dry-matter-content of grassland site-specific. NIRS is used on silage machines and therefore used for harvesting processes. With a deviation of 2% the NIRS-system is very accurate on measuring grass yield. The results are send wireless to a server, and software puts all the different data together in just one yield-overview. The NIRS-system can be installed on a loader wagon.



The benefit of utilization of a sensor-based yield measurement can provide farmers real-time information about their grass quality. Therefore making more accurate decisions to improve both yield and quality of silage to increase production of the dairy herd.

The amount of farms with employees worldwide is growing steadily. On the one hand this is a positive evolution of the dairy farming industry, but this means farmers need to become managers. How can these farm managers be sure all employees perform their work according to the vision of the farm manager?

Writing Standard Operating Procedures (SOPs) for a specific farm has a strong added value. By starting to collate all the activities step-by-step and discuss procedures, more insight into “usual” activities will be created. Standard activities which are normally performed automatic will now be undertaken more consciously. Besides it will be much easier to instruct employees or farm visitors. When SOPs are well integrated into the farm management system it is required for employees to follow up these procedures. When employees are familiar with their tasks the hazard of making mistakes will decrease, and workers can be substituted due when, for example, in the case of sickness. The major advantage for working with SOPs on dairy farms is that routine work will provide a better overview of work and, therefore, better planning. In addition less indirect hours will be incurred, making it more effective. Moreover the SOPs can make more use of the quality of the work. A clear SOP can, therefore, provide greater satisfaction in the long term.

It is difficult to have a clear overview of the grass growth of all the pastures nowadays. The pastures became larger and the weather more instable, so to predict the grass growth and the dry matter content has become more complicated as well as creating an effective planning.

Using the platemeter, as described on the 4D4F website, and performing a FarmWalk weekly gives farmers the opportunity to create awareness about grass growth and grass yield. During a FarmWalk, farmers will walk through the pastures in a V or W-shape with the grass platemeter. The platemater records in 30 measurements the average amount of dry matter per hectare. As a result of performing a FarmWalk regularly farmers can predict grass growth better and create an more effective planning, which can significantly reduce grassland losses and increase the quality of the grass. Using the platemeter is not only interesting for farmers who only mow the grass as a way of harvesting but also for grazing, the use of the platemater is a useful tool. Maybe the added value will even be greater. Especially for farmers who face difficult pasture planning with a significant number of cows per hectare of home ground. Even in this last case the grass intake of the herd can be measured.

The platemeter will help farmers to optimize the utilization of grassland and will be useful by creating an effective planning and insight in the quality of grass. Therewith the cheapest feed type available will not only be cheap but also of higher quality for the cows.

It is very hard and would be time consuming to monitor the standing and lying time as well as the activity of individual cows. This information however can be strongly linked to the health status of each cow individually.

Through using activity sensors, the average lying time for your herd can be established. The concept of daily time budgets can be used to monitor this and allied with activity sensors located on the cows; either foot mounted or neck mounted, an accurate pattern can be established.

Leg mounted sensors however are the only ones that can gather accurate lying time data. It is commonly suggested that cows make more milk when they are lying down as blood flow through the external pudic artery increases by 24-28% when lying compared with standing up. Failure to achieve the required level of rest results in a significant stress response, causing increased lameness risk and vulnerability to health problems.

There are strong links to suggest threshold, and that all cows regardless of yield require a minimum rest period. Once this threshold has been reached a linear relationship may exist with trials showing that after 10 hours rest, every additional hour can generate 1.6 Kg of extra milk without an increase in DMI – simply owing to increased efficiency of milk production.

In addition to an increase in daily yield from appropriate length rest periods, the effect on cow health, a reduction in lameness, reduced cow replacement rates can be a significant economic and welfare driver.

In the past, a herdsman’s notebook sufficed but with data coming from several sources - the parlour, feeding system, third party recorders, veterinary surgeons etc it is no longer adequate – and will lead to possible mistakes. In addition, with several data streams a certain amount of reconciliation will be required to bring all this data together in an accessible form and this is better done automatically on the internet.

Office based desk top computers and even laptop computers are useful on farms but in the cow barn itself best practice would be single entry via an Android or IOS smartphone or similar tablets. Phone network coverage is not essential in that a farm based WIFI system would allow access to the web and “cloud” on which all data will be stored, manipulated and accessed. The WIFI system should be set up to avoid black spots with low conductivity to prevent frustration and allow herdsmen to check data whilst going about their normal work or at location of specific cows. Such a system is available to all farms with access to the internet.

Inspecting a cow directly making breeding judgements, daily management decisions etc whilst accessing full data held on the specific cow allows more accurate and informed decisions.

Single entry systems are likely to be more accurate and reduce the amount of work required – freeing up the herdsman for greater input in critical issues of cow care and stockmanship. The choice of actual phone is personal preference but ensure sufficient processing capacity and memory are available to avoid delay and frustration.

The use of data has been key to effective management of dairy cows for many years but in this time of “big data” and the increase in number of sensors used on farms the increase in amount of data is exponential.

The challenge is to ensure all this data is used to the benefit of the business, the animals and the labour and not simply collected and unused. The data produced by sensors and systems is generally represented in three ways;

• Indices – system gives an overall rating by combining different measures into an index.

• Percentage probability – giving chance/likelihood of an incident/condition occurring.

• Categorise cows – number of cows showing certain trait eg oestrous, mastitis, lameness etc



The challenge to us as dairy animal managers is to combine this data into a means by which it is used – an action list of cows to concentrate on that is often the most applicable but it must be readily understood and communicated to ourselves or our team members, either as hard copies or increasingly through hand held electronic devices – smartphones etc. This combining of the data must be done automatically and not manually as this is ineffective use of skilled labour.



When deciding what system or systems to install on the farm it is crucial that dairy managers critically assess the compatibility of the data streams. Ensure that the decision made suits the farm’s individual needs and that the huge amount of data available is used effectively and not ignored because of data overload or difficult to access daily.

The use of data has been key to effective management of dairy cows for many years but in this time of “big data” and the increase in number of sensors used on farms the increase in amount of data is exponential. The fact that much of the data comes from different source and streams exacerbates the challenges.

The manipulation into an accessible form is both time consuming and complex, and therefore not good use of cow managers time. Use of cloud based system and effective use of the internet based systems is the most appropriate method of manipulation of this data.

This method is the only way in which artificial intelligence and machine learning of individual cow’s behaviour can be accessed which will result in more accurate of alerts and reduce the number of false positives - a challenge at present.

The form in which the final data is presented can be determined in part by the manager’s preferences, but generally comparison with historic data, benchmarking against known KPIs is an appropriate methodology.Cloud based systems will allow real time data to be collected and be shown to advisors with confidence. They will also result in more accuracy and a reduction in work – allowing this to be used in more business-critical areas or simply less hours worked and an economic saving and increased satisfaction.

Goats are seasonal breeders; their breeding cycle is heavily influenced by daylight intensity and length. To maintain as level a milk profile as possible in the “off season” - this tendency must be managed effectively. Naturally goats in Europe will start their oestrus cycles as the day length diminishes in the autumn. In order to get more pregnancies in the 'off' season, we need to influence the natural oestrus cycle of the doe so that she can breed when required by milk profile demands.

The use of remote light sensors and automated lighting systems as described on the 4D4F website are ideal for the management of lighting; optimising breeding opportunities and reducing labour requirements.

The management regime must manipulate the length of time the goats are exposed to light so as to simulate the daylight lengths experienced in the late summer - early autumn period. This requires sufficient light for a minimum of 16 hours per day for a minimum of 45 days, and it is recommended that the light intensity be 200 lux when measured at eye level of the doe. This can be easily verified by installing a free app on your smartphone. The length of light period is then gradually decreased by 1-2 hours per week, until 8-10 hours of light per day is achieved. It is very important that the breeding bucks are exposed to the same light regime so they are equally prepared for the breeding period. Approximately 6-8 weeks after the termination of the light treatment the breeding bucks should be introduced to the does, with fertile oestrus occurring 10-20 days after this.

Milk contracts do stipulate milk profile requirements, and this is a cost effective, non-invasive method of ensuring such commercially important criteria are met.

Feed is the major cost on any dairy goat farm. It accounts for 30-50% of the total cost price. It is clear that farmers should manage this very thoroughly. Nevertheless, in a questionnaire completed by 35 dairy goat farmers, 53.3% of the farmers indicate not to have (enough) insights on economical KPI’s like feed cost, feed efficiency, feed profit, total cost price…

This observation was one of the reasons to establish Farmdesk (www.farmdesk.eu), a spin-off of Wim Govaerts & Co, one of the 4D4F project partners. Farmdesk is an on-line software tool that consists of a Milk, Feed and Economy module. The Economy module allows farmers to calculate the major economical KPI’s on a regular basis, to keep track on it. By collection data from the Milk and Feed module, the calculation is very quick with only a handful of manual inputs.

Many farmers rely for a huge part on ration advice by commercial feed suppliers. Farmers should pay attention that they keep track on the decisions themselves, especially when it comes to ration costs. Feed suppliers have benefits in selling more concentrates, while farmers should make the technical-economical best ration.

Goat milk production results contain an enormous amount of useful data. Based upon production numbers, fat and protein levels, urea levels, cel counts and plate counts (both on global farm level as on individual animal level), farmers should continuously make decisions. However, many dairy goat farmers admit not to have enough insights on the data.

This observation was the reason to establish Farmdesk (www.farmdesk.eu), a spin-off of Wim Govaerts & Co, one of the 4D4F project partners. Farmdesk is an on-line software tool that consists of a Milk, Feed and Economy module. The Milk module collects all milk data through automatic data connections. Farmdesk visualizes the data and gives automatic attentions by smart algorithms.

Farmdesk and its mobile app give automatic push notifications to farmers in case of attentions. In this way, farmers keep track on this, also in busy periods e.g. in field seasons. On a technical-economical level, preventing heat stress issues (observable by low fat/protein ratios) by adding rumen buffers, preventing ketosis (observable by high fat/protein ratios) by adding more energy or maximizing milk production by monitoring the urea levels, are ways to prevent economical losses that can quickly rise till 10000 EUR on an average dairy goat farm.

Feed management is one of the major concerns on any dairy goat farm. On a technical level, rationing has a huge impact on milk production. It accounts for 40-70% of the milk yield. In a questionnaire completed by 35 dairy goat farmers, only 9.7% of the farmers indicate to calculate the rations themselves, while 67.7% indicates the willing to do it themselves.

This observation was one of the reasons to establish Farmdesk (www.farmdesk.eu), a spin-off of Wim Govaerts & Co, one of the 4D4F project partners. Farmdesk is an on-line software tool that consists of a Milk, Feed and Economy module. The Feed module allows farmers to administer their feeds, and to calculate rations in a very graphical, user-friendly way.

By presenting the ration in a graphical way, farmers pick up insights much quicker. They learn to know the various fodders. Also, Farmdesk incorporates animal signals (body condition score, manure consistency, milk data) into account to give hands-on advice to adjust rations accordingly. Next to this, a farmer can connect his/her farm to his/her preferred advisor. In this way, the communication process between farmer and advisor is much more efficient. The farmer pays less on consultancy hours, the advisor gains a lot of time not searching/mailing/calling for data, he sees and communicates it on-line.

Having healthy and productive replacement stock is vital for the future successes of any dairy herd. Therefore, calf management is a very important process for both short- and long-term benefits. Managing any health and disease risk is a vital part of this; to help benefit in terms of fertility, health and production in the future. Fever tags are a temperature monitoring ear tag that measures the core temperature of the calf from a probe in the ear canal every 15 minutes. Once the threshold of temperature is breached over a certain time period, this then signals an alert and the ear tag will flash. This allows any farmer or stock person to have an early indication of disease or health problems up to 72 hours before any visual signs can be seen. As a result of this, any at risk calves can be treated before any further health issues occur, such as reduced appetite or dehydration. As well as improve health and management of the calves, farmers can monitor the spread of a disease and also make better use of preventative treatments and anti-inflammatory medication, thus reducing antibiotic use.

Farms are becoming more efficient at identifying lame animals. There are however still costs associated with finding the specific causes. These costs can be through misinterpretation, leading to wasting administered antibiotics, as well as labour and time cost. How thermal imaging can be used is through measuring infrared radiation. Infrared radiation is directly affected by temperature, as with a higher temperature more radiation is emitted. With using thermal imaging cameras, specific causes of lameness can be found from areas of higher temperature. With this equipment, better diagnosis can be made for cause of lameness. For example, identifying which specific claw and area of the foot is the cause, or whether the problem is in the foot or a different area of the animal. The benefits from this include early detection and confidence when choosing course of treatment; resulting in less antibiotics being used and better welfare for the animals affected. Through early detection, there will be less severe cases of lameness, contributing to improved overall health of the dairy herd.

Lameness is one of the biggest and most widespread problems experienced by dairy farms. In any degree of severity, having a high lameness percentage in your herd can be very costly. The costs themselves can affect indirectly in terms of production, antibiotic use and labour; and also directly impacting cow health and welfare. Identifying lame cows by eye is still a popular method currently, however in response to increasing average herd size it will become important to be more efficient in achieving earlier lameness detection. The CowAlert system allows automatic lameness detection with the use of a pedometer containing an accelerometer. This measures activity levels, step count and standing and lying time; every day, 24hrs a day. This allows earlier and more accurate lameness detection, picking up problems that may not have been picked up by the human eye. With the use of CowAlert, farmers can better manage lameness problems sooner, allowing closer monitoring or in some cases antibiotic treatment. Overall the benefits are reduced antibiotic use through earlier detection, better herd health and welfare, reduced labour costs as well greater milk production potential.

Ketosis is often caused by the fact that after calving a cow doesn’t consume enough energy to meet her needs. A cow can get very sick from ketosis, which causes her to lay down a lot and eat less which increases the effect of ketosis. A cow that has a BCS of >3.75 at calving has a greater risk at ketosis. A high BCS at calving occurs when an energetic ration is fed in the dry period. Dried off cows that eat a lot of energy before calving have a lower rumen fill, because they eat less dry matter. Which results in a lower intake capacity after calving. The BCS camera enables the farmer to monitor his cows closely in vulnerable periods and adjust the feed more specifically when the BCS of a cow is increasing too much. As discussed above a cow can be too fat at calving but it can also have a BCS of <3.25 which means there was too little energy provided in late lactation and dry period. This increases the risk at reproduction problems and a lower milk production during the lactation. Thereby high milk producers might drop below a BCS of 2.75 which can also cause problems regarding the reproduction. And lastly when a cow has a BCS > 3.75 at dry off this might cause problems at calving and reproduction in the next lactation. If the BCS is scored regularly using the BCS camera, a more specific feeding management can be applied on the farm. Group averages can be assessed and feed can be adjusted according to the BCS. Thereby, individual cows are monitored. While improving the feeding strategy to prevent ketosis the farmer might also see an improvement in reproduction, higher milk production and a better health of the herd. The camera can be placed at a selection gate after the milking parlour or robot and therefore is a stress-free method for determining BCS.

Heat stress has long been recognised as a problem in Mediterranean countries, but global warming has increased its importance in all European dairy farms.. The number of days with temperatures over 30 C in southern UK has doubled in last 20 years. Heat stress and subclinical effects can result in reduced output and suboptimal herd fertility, and can be seen with ambient temperatures as low as 20 C. Increased natural or forced ventilation can be used to mitigate the challenge of hot summers but misters and even refrigeration may be necessary. Rumination sensors can be used to test the efficacy of this action on individual cows, and even highlight a problem in the first place. The management of this ventilation however must be considered and decisions required for its effective management should not be made from subjective human criteria but from accurate data derived from embedded sensors throughout the barn which integrate the data with humidity data – and produce a Temperature Humidity Index. Such systems then activate fans and or inlet manifolds so the temperature is kept close to optimum for health and milk production. Such systems can be retro fitted to existing buildings or incorporated in new cow barns .As temperatures are forecast to rise, the modern cow manager will have the ability and technology to incorporate accurate sensors in the cow barns to ensure cows will not suffer heat stress and have health and production compromised.

Critical to the effective management of data is the level of false positive alerts resulting in wasted time, poor decision making and a permanent loss in the level of confidence in the system. Staff facing such inaccuracies will become sceptical of the system and adopt methods of working that do not fully utilise the benefits that data can bring to the management of the dairy cow. “Unpolluted” data is core to a successful business – particularly where farms rely on analysis of large quantities of data to make crucial decisions and ultimately survive.

The solution to high levels of false alerts or indeed too few alerts is to access the system routinely and link all alerts with specific farm based physical cow knowledge, using the lists produced to check specific animals. Adding cow specific behaviour into data enables the system to incorporate such animal behaviour into its alerts.

In addition to the introduction of cow specific behaviours into the data set, the alerts thresholds can be altered to ensure that alerts level is correct, not too many and not too few. A continual process of feedback is required to ensure that action lists produced are legitimate – data cannot and should not be used in isolation but worked with alongside and reacting with physical data/behaviour which is observed on the farm. This process can then lead to the formulation of farm specific operating procedures.

Small herds have specific problems with heat detection. Numerous factors affect the expression of heat including, housing arrangement, floor surface, feet and leg problems and status of herd mates. The number of mounts a cow receives increases with the number of cows that are in heat simultaneously up to about 3–4 cows in heat. Cows that are themselves in heat, coming into heat or were recently in heat are most likely to mount a cow that is in heat. Cows that are pregnant show less interest in mounting other cows that are in heat. In smaller herds, with lower numbers of open cows, the likelihood of more than one cow being in heat on any given day becomes less, consequently, making heat detection more difficult. In addition to this lack of mounting mates, small herds face specific problems which can lead to extended calving intervals, including labour availability, possible housing deficiencies and shorter oestrous caused by higher yields. The 4D4F website details the different technologies available to cow managers. Sensors can monitor cows 24/7 and produce daily action lists for cow managers informing them of cow identity and optimum time for insemination. With a benefit of 4 euros per day from reducing calving interval - a 100 cow herd reducing calving interval from 420 to 400 days with the use of sensors monitoring 24/7 will see an 8,000 euro increase in profits and will in addition benefit from lower veterinary treatments costs and lower labour requirements.

Critical to the effective management of data is the level of false positive alerts resulting in wasted time, poor decision making and a permanent loss in the level of confidence in the system. Staff facing such inaccuracies will become sceptical of the system and adopt methods of working that do not fully utilise the benefits that data can bring to the management of the dairy cow. “Unpolluted” data is core to a successful business – particularly where farms rely on analysis of large quantities of data to make crucial decisions and ultimately survive.

The solution to high levels of false alerts or indeed too few alerts is to access the system routinely and link all alerts with specific farm based physical cow knowledge, using the lists produced to check specific animals. Adding cow specific behaviour into data enables the system to incorporate such animal behaviour into its alerts.

In addition to the introduction of cow specific behaviours into the data set, the alerts thresholds can be altered to ensure that alerts level is correct, not too many and not too few. A continual process of feedback is required to ensure that action lists produced are legitimate – data cannot and should not be used in isolation but worked with alongside and reacting with physical data/behaviour which is observed on the farm. This process can then lead to the formulation of farm specific operating procedures.

Small herds have specific problems with heat detection. Numerous factors affect the expression of heat including, housing arrangement, floor surface, feet and leg problems and status of herd mates. The number of mounts a cow receives increases with the number of cows that are in heat simultaneously up to about 3–4 cows in heat. Cows that are themselves in heat, coming into heat or were recently in heat are most likely to mount a cow that is in heat. Cows that are pregnant show less interest in mounting other cows that are in heat. In smaller herds, with lower numbers of open cows, the likelihood of more than one cow being in heat on any given day becomes less, consequently, making heat detection more difficult. In addition to this lack of mounting mates, small herds face specific problems which can lead to extended calving intervals, including labour availability, possible housing deficiencies and shorter oestrous caused by higher yields. The 4D4F website details the different technologies available to cow managers. Sensors can monitor cows 24/7 and produce daily action lists for cow managers informing them of cow identity and optimum time for insemination. With a benefit of 4 euros per day from reducing calving interval - a 100 cow herd reducing calving interval from 420 to 400 days with the use of sensors monitoring 24/7 will see an 8,000 euro increase in profits and will in addition benefit from lower veterinary treatments costs and lower labour requirements.

Heat stress has long been recognised as a problem in Mediterranean countries, but global warming has increased its importance in all European dairy farms.. The number of days with temperatures over 30 C in southern UK has doubled in last 20 years. Heat stress and subclinical effects can result in reduced output and suboptimal herd fertility, and can be seen with ambient temperatures as low as 20 C. Increased natural or forced ventilation can be used to mitigate the challenge of hot summers but misters and even refrigeration may be necessary. Rumination sensors can be used to test the efficacy of this action on individual cows, and even highlight a problem in the first place. The management of this ventilation however must be considered and decisions required for its effective management should not be made from subjective human criteria but from accurate data derived from embedded sensors throughout the barn which integrate the data with humidity data – and produce a Temperature Humidity Index. Such systems then activate fans and or inlet manifolds so the temperature is kept close to optimum for health and milk production. Such systems can be retro fitted to existing buildings or incorporated in new cow barns .As temperatures are forecast to rise, the modern cow manager will have the ability and technology to incorporate accurate sensors in the cow barns to ensure cows will not suffer heat stress and have health and production compromised.